The goal of this thesis is to realize some critical techniques required in the cognitive radio mesh network. Cognitive radio (CR) is a new technology introduced to improve the frequency efficiency of current wireless systems. A mesh network, which leverages the advantages of an infrastructure network and a pure ad-hoc network, is a network topology highly suitable for the CR networks. CR users need to adapt to their ambient wireless environment and automatically select their own effective transceiver mode. The thesis focuses on the realization of a physical layer protocol, dynamic frequency-selection algorithm, mesh network signalling method, and transmit-power control algorithm in CR mesh networks. A new dynamic frequency selection algorithm is introduced in CR network, including estimation of the primary users' traffic statistics as well as adaptation to the local background noise interference. Through continuous observation, analysis, and adaptation to the time-varying environment, CR is able to select the candidate frequency bands to satisfy a user's rate and power requirements without causing collision to the primary users in those bands. As a generic model, a Gaussian mixture model, is selected for characterizing the statistics of the traffic environment. The transmit-power control in the CR mesh network is based on an iterative water-filling algorithm, rooted in information theory. The iterative water-filling algorithm is modified to suit the requirement of CR mesh networks. It works well for setting a suitable transmit-power level, sub-channels, and sub-carriers for the nodes in a CR mesh network. To support current systems, experiments are presented on the wireless mesh network, which are based on the orthogonal frequency-division multiple-access (OFDMA) mode of the High-speed Unlicensed Wireless Metropolitan Area Network (Wireless HUMAN). In the mesh network, the multiple access control(MAC) layer signals including the distributed network configuration, network entry, and network scheduling signals, are introduced. They are transmitted through a control channel in a cooperative mechanism. The physical layer(PHY) of OFDMA is also discussed in the thesis. / Thesis / Master of Applied Science (MASc)